Indian Journal of Pathology and Microbiology

: 2019  |  Volume : 62  |  Issue : 3  |  Page : 368--374

Revisiting metastatic central nervous system tumors with unknown primary using clinicopathological findings: A single neurosciences institutional study

Anshu Gupta1, Sujata Chaturvedi1, Deepak Jha2, Monali Chaturvedi3,  
1 Department of Pathology, Institute of Human Behavior and Allied Sciences, Delhi, India
2 Department of Neurosurgery, Institute of Human Behavior and Allied Sciences, Delhi, India
3 Department of Radiology, Institute of Human Behavior and Allied Sciences, Delhi, India

Correspondence Address:
Anshu Gupta
1408/13, Opposite Model School, Civil Road, Rohtak - 124 001, Haryana


Background: Metastatic tumors are the most common central nervous system (CNS) tumors wherein the primary site remains unknown in most of the cases. Aim: The study was carried out to evaluate metastatic CNS tumors with unknown primary by using simplified diagnostic (clinico-histopathologic) approach. Material and Methods: A 2 years study was conducted on 32 cases of CNS metastases having unknown primary tumors in a neurosciences institute. Statistical Analysis: All the results were prepared using software version of SPSS 22. Results: The most common metastatic site found in brain was cerebrum (59.3%) [frontal > frontoparietal > parieto-occipital > temporal] [left cerebrum > right cerebrum], followed by cerebellum (12.5%), spinal cord (9.3%), and leptomeninges (3.12%). Most of the metastatic tumors presented as ill-defined (34%) rather than well-defined (22%) lesions with ring enhancement seen only in 16% of the cases on magnetic resonance imaging (MRI).On histopathology findings with targeted immunohistochemistry, most common histological tumor type identified irrespective of site was adenocarcinoma (68.7%), followed by squamous cell carcinoma (15.6%) and poorly differentiated carcinoma (12.5%). Only one case of lymphoma was reported. Corroborating all the above findings along with clinical history and other relevant investigations, primary sites could be detected in 23 cases (71.8%).The most common primary site deduced was lungs (39.1%), followed by thyroid (17.3%), breast in females (13.0%), gastrointestinal tract (8.6%), and prostate in males (4.3%). Only in nine cases (28.1%) with mainly poorly differentiated histopathological type, primary site remained unknown. Conclusion: Detection of the primary site in metastatic CNS tumors is possible by adopting this simple and effective diagnostic approach at centers/hospitals having cost and other constraints.

How to cite this article:
Gupta A, Chaturvedi S, Jha D, Chaturvedi M. Revisiting metastatic central nervous system tumors with unknown primary using clinicopathological findings: A single neurosciences institutional study.Indian J Pathol Microbiol 2019;62:368-374

How to cite this URL:
Gupta A, Chaturvedi S, Jha D, Chaturvedi M. Revisiting metastatic central nervous system tumors with unknown primary using clinicopathological findings: A single neurosciences institutional study. Indian J Pathol Microbiol [serial online] 2019 [cited 2019 Aug 21 ];62:368-374
Available from:

Full Text


Metastases are the most common central nervous system (CNS) tumors. Metastatic CNS tumors are mostly symptomatic at the time of diagnosis and present with headache, seizures, localized motor deficits, and dysphagia.[1] Brain metastases contribute significantly to morbidity and mortality of patients with primary solid tumors despite the fact that brain has a very low rate of metastases.[2] Exact incidence is unknown, however, the incidence of brain metastases according to population based studies range from 8.3 to 14.3 per 100,000 people.[3] Common primary sites for brain metastasis are lung (20%), kidney (6.8%), skin (6.5%), breast (5%), and gastrointestinal tract (GIT, 1.8%) which account for majority of brain metastases.[4] Brain metastasis usually occurs late during the course of malignancy. Interval between diagnoses of primary tumor and brain metastases showed significant variation ranging from 4 months to 37 months according to a study.[5]

Sometimes, first signs and symptoms of malignancy appear after brain metastasis that require clinical, radiological, and pathological work-up for an unknown primary. Such cases account for 16% CNS metastases and majority of them turn out to be of pulmonary origin.[6] A hospital-based study was carried out to analyze metastatic CNS tumors and their primary sites of origin, following a systematic approach correlating clinical, radiological, and histopathological findings along with relevant immunohistochemistry.

 Material and Methods

This is a 2 years descriptive and retrospective study conducted in a tertiary neurosciences institute from January 2014 to December 2015. The clinical information was retrieved from medical records and variables noted were patient's names, age, sex, registration numbers, clinical history, radiological findings, and histopathologic details with tumor type diagnosed. Metastatic tumors in which primary site could not be found at first investigations were analyzed. For this, further clinical data was extracted for likely primary sites and correlated with radiological features and routine hematoxylin and eosin (H and E) stained sections findings. Depending upon the histopathological findings, only relevant immunohistochemical stains were done on formalin-fixed paraffin embedded tissues. Various immunohistochemistry (IHC) markers used for different cases were glial fibrillary acidic protein (GFAP), cytokeratin (CK), epithelial membrane antigen (EMA), leukocytes common antigen (LCA), synaptophysin, chromogranin, neuron-specific enolase (NSE), and thyroid transcription factor-1(TTF-1). GFAP and epithelial markers (EMA, CK) were used in those cases where there was difficulty in differentiating primary glial tumor from secondary tumor. Synaptophysin and chromogranin were used for neuronal differentiation. LCA was done to rule out tumors of lymphoid origin whereas progesterone receptor (PR) was used to confirm breast carcinoma. TTF-1 was used to detect primary site in cases of brain metastases with unknown primary site. Immunostaining was interpreted in every case in light of clinical and radiological findings.


An ethical approval was taken to conduct this study.


Data was analyzed using Statistical Package for the Social Science version 22.0 (SPSS Inc; Chicago II, USA). The results were expressed as mean for quantitative variables and as percentages for qualitative variables.


During this 2 years period, a total of 450 CNS tumors were diagnosed histopathologically including 345 (76.6%) primary and 105 (23.3%) secondary tumors. Out of these 105 cases, 32 (30.4%) were of metastatic CNS tumors with unknown primary in which patients had primarily CNS manifestations only such as headache, seizures, muscle weakness, etc. Myeloproliferative tumors were not included in the study. Ages of all patients ranged from 18 to 80 years. The mean age was 65 years. Maximum numbers of cases were observed in fifth and sixth decades. Gender wise, male:female ratio was 1.4:1 (males-19, females-13).

Depending upon the number of metastatic lesions, the distribution was: single lesion (including solitary) in 81.2% cases and multiple lesions in 18.7% cases.

The anatomical sites of metastatic lesions were: cerebrum-59.3%, cerebellum-12.5%, spinal cord-9.3%, leptomeninges only-3.12%, and multiple sites-15.6%. In the cerebrum, most common site involved was frontal-31.5% followed by frontoparietal-2l% and parieto-occipital-21% and then occipital-10.5%, temporal-10.5%, and parietal-5.2%. Left cerebral hemisphere was more commonly involved than right hemisphere.

The common symptoms noted were headache-71.8%, followed by abnormal behavior-25%, vomiting-12.5%, seizures-9.3%, and diplopia/blurring of vision-9.3%. MRI findings have been summarized in [Figure 1].{Figure 1}

Routine H and E stained sections were evaluated and various broad histopathological diagnoses were made as shown in [Table 1]. After applying selective immunohistochemical stains and correlating with clinicoradiologic findings, we were able to detect primary sites in 71.87% (n = 23) cases as illustrated in [Table 2]. In rest of the cases, that is, 28.13% (n = 09) cases, primary sites remained unknown.{Table 1}{Table 2}


Most common tumors in brain are metastatic tumors. Brain metastasis is evident in 20--40% of all cancer patients and usually occurs late during the course of primary malignancy. Overall incidence of brain metastases is 10--12 in 10,000 people but probably that underestimates the true incidence due to nonavailability of adequate data.[4] 60--80% of metastatic brain tumors originate from primary cancers of lung, breast, skin (melanoma), kidney, and GIT.[7] The large majority of metastases arise from carcinomas and only 3% of patients with sarcoma develop metastases as reported by a study.[8] 16% of carcinomas remain unrecognized before metastasizing to brain.[5] In our study, 23.3% cases were of metastatic CNS tumors which was lower than the figures reported in other studies,[1],[2],[3],[4] which could be due to many reasons. This study was purely based on the data of a tertiary neurosciences institute with no radiotherapy services. So, those patients who required radiotherapy were referred to multispeciality hospitals having both neurosurgery and radiotherapy disciplines. Only surgical pathology material was considered for the study as no autopsy material was available. Sometimes, patients with metastatic CNS tumors did not turn up for work-up and treatment due to their dismal state leading to lower proportion of secondary tumors in our institute.

Majority of neoplasms involve CNS by hematogenous route or by direct extension from primary solid tumor.[9] Majority of CNS metastases reach the brain either through arterial circulation or by Batson venous plexus. Most of the brain metastases are parenchymal; however, metastases to leptomeninges, dura, and skull are not uncommon. Within brain, corticomedullary junction and water shed zone, mainly supplied by middle cerebral artery are most commonly involved. Metastasis is most commonly seen in cerebral hemispheres (80%), cerebellum (10-15%), brain stem (2-3%), and less commonly in spinal cord, dura, leptomeninges, pituitary and choroid plexus, depending upon the amount of blood flow to each area. Cerebellum involvement is seen commonly in colorectal, uterine, and renal carcinomas. The spinal epidural metastases most commonly occur from cancers of prostate, breast, lung, and kidney.[10] In the cerebrum, frontoparietal area is most commonly involved.[2] Our study showed most common involvement of cerebrum (59.3%), followed by cerebellum (12.5%), spinal cord (9.3%), and leptomeninges only (3.12%). In the cerebellum, metastases noticed were from breast and GIT. Carcinomas from breast and cervix involved spinal cord. Multiple sites involvement was seen in 15.6% cases.

The signs and symptoms of brain metastases are usually due to increased intracranial pressure or local effect of tumor on adjacent brain tissue. Although the predominant presenting symptom is headache but the patient may have motor or language deficits and seizures. Approximately 10% of patients are asymptomatic and require imaging studies for diagnosis.[11] In our study, we noted headache as the most common symptom. Other symptoms observed were abnormal behavior, vomiting, seizures, and diplopia/blurring of vision.

It is important to follow a systematic approach while evaluating metastatic neoplasms. One should consider factors like age, sex, and location and then factors that help in narrowing down the differential diagnosis such as radiological findings, gross pathology, histopathological features, followed by targeted use of IHC and molecular studies.[12]

The preferred modality for detecting brain lesions is MRI or MRI with contrast; however, computerized tomography (CT) imaging or contrast CT is often used in cases of nonavailability or nonsuitability to the patient.[11] On imaging studies, the metastasis could appear as solitary (one CNS lesion without other systemic metastases), single (one CNS lesion with other systemic metastases), or multiple lesions.[13] On MRI, metastatic lesions appear as well-defined, ring enhancing lesions surrounded by edema. Most of them show mild T1 hypointensity with T2 hyperintensity. Metastases may be associated with hemorrhage or necrosis. Some malignancies are more susceptible to hemorrhage such as melanoma, choriocarcinoma, renal cell carcinoma, thyroid and lung cancer. Hemorrhagic metastases may also demonstrate T1 signal hyperintensity. Cystic necrosis demonstrates high signal on fluid-attenuated inversion recovery (FLAIR) sequence (which suppresses signal intensity that is cerebral spinal fluid [CSF]-like).[14],[15] On contrast studies, lesions may show ring, nodular, solid enhancement, or delayed enhancement. The findings of high grade glioma and metastases are similar on MRI, making it difficult to distinguish between them. Prevalence of multiple lesions is 55% in brain metastases, whereas it is 23% in gliomas. Thus, multiple brain lesions are more suspicious for brain metastasis.[11] Lung cancer and melanoma are more likely to be associated with multiple lesions whereas breast, thyroid, renal, and colorectal carcinomas present with a solitary metastatic lesions.[9],[10] In the present study, on MRI, ill-defined lesions were noted in 34.3%, whereas well-defined lesions were seen in 21.8% of the cases. Ring enhancement was seen in 15.6% cases. Single lesion (including solitary) was observed in 81.2% cases and multiple lesions in 18.7% cases, as shown in [Figure 2] (2.1-2.5). These findings are in concordance with radiological findings reported in literature.[14],[15]{Figure 2}

Diagnosis of brain lesions on imaging must be confirmed by pathology using biopsy of the tumor tissue. Grossly, metastatic lesions are usually sharply demarcated, spherical masses that do not infiltrate the surrounding brain parenchyma. Large tumors tend to have necrosis and hemorrhage. Metastatic lesions may show characteristic features of primary tumor such as shiny mucoid appearance in mucin secreting adenocarcinomas and gray to black pigment in melanoma.[16]

Microscopically, evaluate the tumor for the presence of squamous cells, clear cells or mucin secreting cells and architectural patterns such as glandular, papillary, solid nests, trabeculae, comedo pattern or presence of colloid material or dirty necrosis. In adenocarcinomas, it is advisable to look for specific features. Presence of small glands lined by tumor cells having prominent nucleoli and a vacuolated cytoplasm favors a metastatic prostatic adenocarcinoma or presence of large glands, papillary structures, or mucin secretion supports the diagnosis of metastatic lung or pancreatic carcinoma. Presence of solid nests with central area of necrosis suggestive of comedo pattern, composed of tumor cells with mild nuclear pleomorphism or trabecular arrangement strongly suggest the possibility of breast origin, particularly in a female patient. Glandular spaces lined by psuedostratified cells with cigar type nuclei and extensive necrosis with karyorrhectic debris suggest colonic or rectal adenocarcinoma. Colloid within follicles is characteristic of metastatic thyroid carcinoma and a papillary morphology is suggestive of neoplasms of thyroid, gynecologic, or pulmonary origin. Diagnosis of metastatic lung cancer would become the most likely in presence of squamous cells.[12],[16] Recognition of neuroendocrine morphologic features in a metastatic neoplasm suggest the diagnoses of typical and atypical carcinoid tumors and high-grade neuroendocrine carcinomas such as small cell carcinomas. The diagnosis of metastatic small cell carcinoma strongly suggest the possibility of a pulmonary primary lesion although this morphology can develop in a variety of other primary sites such as upper air ways, gynecologic tract, urinary tract, and others.[17] Cells that exhibit a clear or glycogenated cytoplasm suggest primary from two sites, kidney and gynecologic tract, including ovary and endometrium.[18] Germ cell tumors are uncommon neoplasms that usually occur in pediatric or young adults and primary lesion in the gonads and mediastinum needs to be excluded.[19] Melanomas exhibit diverse histomorphological features.[20]

In the lung, the incidence of metastases also varies with the histologic subtype of the tumor. Non-small cell lung carcinomas are more likely to metastasize than small cell carcinomas and among the former, adenocarcinomas are more likely to metastasize than squamous and large cell carcinomas.[21],[22] In our study, most common histopathological diagnoses observed were adenocarcinomas including papillary adenocarcinomas in 22 cases (68.7%), followed by five cases (15.6%) of squamous cell carcinomas and four cases (12.5%) of poorly differentiated carcinomas, shown in [Figure 3] (3.1-3.4). Only one case (3.12%) of non-Hodgkin lymphoma, diffuse large cell type, was seen.{Figure 3}

The first step in the diagnosis of a metastatic brain tumor is to exclude a primary CNS tumor and the second step is to identify primary tumor and the site. When the metastatic tumor is well differentiated, diagnosis is rarely a problem. In case of a poorly differentiated neoplasm, especially when solitary, a high grade glioma should always be considered in differential diagnosis.[11] High grade glioma shows an invasive border but limited and perivascular invasion can also be seen in metastatic small cell lung carcinoma, melanoma, and lymphoma. In such cases, ancillary IHC studies are required for further evaluation.[23] Negative staining with CK, EMA in addition to positive GFAP favors diagnosis of high grade glioma. IHC becomes even more important when morphology is not enough to make diagnosis and in evaluation of metastatic neoplasm of unknown primary (MNUP). MNUP accounts for nearly 16% of the cases.[22] In majority of unknown primary with CNS metastasis, lung is the primary site. A basic IHC panel is a widely accepted initial step in difficult and MNUP cases such as GFAP (gliomas), CK, and EMA (carcinomas), S-100 (melanoma, glioma) synaptophysin, chromogranin and NSE (neuroendocrine tumors), LCA (lymphoma) HMB-45 (melanoma), estrogen receptor and progesterone receptor (breast carcinoma) and prostate-specific antigen (prostatic carcinoma).[24],[25] Armamentarium of more specific IHC markers is often applied in the second round staining based on initial findings and suspected primary, as shown in [Figure 4]. Numerous literature data are available on specific IHC staining to diagnose primary site of brain metastases. Out of all, thyroid transcription factor-1 (TTF-1) is an important marker to identify lung as the primary site of origin. Normal brain tissue and astrocytic tumors do not stain with TTF-1. This assumes decisive importance in MNUP, taking into consideration the fact that lung is the most common primary in cases of MNUP. If TTF-1 is negative, one can rule out other less common primary tumors. Diagnosis of ovarian carcinoma is made on CA 125 positivity, carcinoembryogenic antigen (CEA), and vimentin negativity. Positive human epidermal growth factor receptor-2 (HER2), GCDFP-15, and CK7 positivity favors diagnosis of breast carcinoma. Colonic carcinoma is diagnosed by positivity for CK7, CK 20 and CEA and by TTF-1 negativity.[25],[26] The major drawback is cost factor that acts as a deterrent in implementation of all these markers, especially in developing countries. In our study, we applied first core antibodies along with TTF-1 and primary sites could be detected in 71.8% of the cases. The most common primary site diagnosed was lung followed by thyroid [Figure 3].5] and breast. Only one case each was noted from colorectum, oesophagus, prostate, buccal mucosa, and ear as primary site. In 28.13% (9) cases, primary sites remained unknown despite all efforts.{Figure 4}

Few previous Indian studies on immunohistochemical analysis of CNS metastases have also reported similar findings. Singhet al. conducted a 4 year study on 69 cases of metastatic brain neoplasms and found lung cancer to be the most common primary malignancy followed by breast, colorectal and skin cancers (melanoma).[27] Sharmaet al. analyzed 74 cases of brain metastases from an unknown primary for a period of 5 years, in which IHC was applied. They concluded that out of 74 cases, the primary site could be detected in 86% cases and the most common was lung followed by female genital tract and breast. The primary site could not be diagnosed in 14% cases even with IHC.[26] Study by Patnayaket al. in a tertiary care center in South India reported that primary site remained unknown in 47% of cases with metastatic adenocarcinoma in brain. The lung was the most common site of metastatic adenocarcinoma followed by thyroid and prostate. Metastatic squamous cell carcinoma was the next followed by metastatic melanoma of the skin, metastatic germ cell tumor, and metastatic neuroendocrine carcinoma of lung. 10% of CNS metastasis was undifferentiated.[25] However, in our study, CNS metastasis from thyroid was found to be more frequent than metastasis from breast, in contrast to the existing literature.

Main limitations of our study were small sample size and availability of only limited panel of antibodies for immunohistochemical staining due to cost constraint and lesser biopsies available. A more elaborative and larger study would be taken up for MNUP later on by applying complete panel of antibodies to detect primary site.

To conclude, evaluation of CNS metastatic neoplasms of unknown primary is a complicated process, requiring clinical history, imaging studies, and knowledge of different tumor types on histopathology. By all these, one can narrow down the likely primary site of origin, even with limited use of IHC instead of adopting the current trend of overly relying on immunohistochemical and molecular tests to identify the site of origin of metastatic neoplasms.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Kamar FG, Posner JB. Brain metastases. Semin Neurol 2010;30:217-35.
2Subramanian A, Harris A, Piggott K, Shieff C, Bradford R. Metastasis to and from the central nervous system: The 'relatively protected site'. Lancet Oncol 2002;3:498-507.
3Nayak L, Lee EQ, Wen PY. Epidemiology of brain metastases. Curr Oncol Rep 2012;14:48-54.
4Barnholtz-Sloan JS, Sloan AE, Davis FG, Vigneau FD, Lai P, Sawaya RE. Incidence proportions of brain metastases in patients diagnosed (1973 to 2001) in the Metropolitan Detroit Cancer Surveillance System. J Clin Oncol 2004;22:2865-72.
5Lagerwaard FJ, Levendag PC, Nowak PJ, Eijkenboom WM, Hanssens PE, Schmitz PI. Identification of prognostic factors in patients with brain metastases: A review of 1292 patients. Int J Radiat Oncol Biol Phys 1999;43:795-803.
6Mavrakis AN, Halpern EF, Barker FG, Gonzalez RG, Henson JW. Diagnostic evaluation of patients with a brain mass as the presenting manifestation of cancer. Neurology 2005;65:908-11.
7Posner JB. Management of brain metastases. Rev Neurol 1992;148:477-87.
8Chu P, Wu E, Weiss LM. Cytokeratin 7 and cytokeratin 20 expression in epithelial neoplasms: A survey of 435 cases. Mod Pathol 2000;13:962-72.
9Nussbaum ES, Djalilian HR, Cho KH, Hall WA. Brain metastases. Histology, multiplicity, surgery, and survival. Cancer 1996;78:1781-8.
10Delattre JY, Krol G, Thaler HT, Posner JB. Distribution of brain metastases. Arch Neurol 1988;45:741-4.
11Campos S, Davey P, Hird A, Pressnail B, Bilbao J, Aviv RI, et al. Brain metastasis from an unknown primary, or primary brain tumor? A diagnostic dilemma. Curr Oncol 2009;16:62-6.
12Marchevsky AM, Gupta R, Balzer B. Diagnosis of metastatic neoplasms: A clinicopathologic and morphologic approach. Arch Pathol Lab Med 2010;134:194-206.
13Stark AM, Stohring C, Hedderich J, Held-Feindt J, Mehdorn HM. Surgical treatment for brain metastases: Prognostic factors and survival in 309 patients with regard to patient age. J Clin Neurosci 2011;18:34-8.
14Fink KR, Fink JR. Imaging of brain metastases. Surg Neurol Int 2013;4:209-19.
15Lingelli A, Khandji AG. Review of imaging techniques in the diagnosis and management of brain metastases. Neurosurg Clin 2011;22:15-25.
16Wronski M, Arbit E. Surgical treatment of brain metastases from melanoma: A retrospective study of 91 patients. J Neurosurg 2000;93:9-18.
17Blaszyk H, Hartmann A, Bjornsson J. Cancer of unknown primary: Clinicopathologic correlations. APMIS 2003;111:1089-94.
18Tang LH, Shia J, Soslow RA, Dhall D, Wong WD, O'Reilly E, et al. Pathologic classification and clinical behavior of the spectrum of goblet cell carcinoid tumors of the appendix. Am J Surg Pathol 2008;32:1429-43.
19Rosenblum MK. Central nervous system. In: Rosai J, editor. Rosai and Ackerman's Surgical Pathology. 10th ed. Missouri: Mosby; 2012. p. 2307-440.
20Berney DM, Warren AY, Verma M, Kudahetti S, Robson JM, Williams MW, et al. Malignant germ cell tumours in the elderly: A histopathological review of 50 cases in men aged 60 years or over. Mod Pathol 2008;21:54-9.
21Miller AJ, Mihm MC. Melanoma. N Engl J Med 2006;355:51-65.
22Hubbs JL, Boyd JA, Hollis D, Chino JP, Saynak M, Kelsey CR. Factors associated with the development of brain metastases: Analysis of 975 patients with early stage non small cell lung cancer. Cancer 2010;116:5038-46.
23Takei H, Rouah E, Ishida Y. Brain metastasis: Clinical characteristics, pathological findings and molecular subtyping for therapeutic implications. Brain Tumor Pathol 2016;33:1-12.
24Pekmezci M, Perry A. Neuropathology of brain metastases. Surg Neurol Int 2013;4:245-55.
25Patnayak R, Jena A, Vijaylaxmi B, Lakshmi AY, Prasad BCM, Chowhan AK, et al. Metastasis in central nervous system: Clinicopathologic study with review of literature in a tertiary care center in South India. South. Asian J Cancer 2013;2:245-9.
26Sharma P, Trivedi P, Shah MJ. Evaluation of central nervous system metastases with immunohistochemistry correlation. Indian J Pathol Microbiol 2014;57:376-9.
27Singh S, Amirtham U, Premalata CS, Lakshmaiah KC, Viswanath L, Kumar RV. Spectrum of metastatic neoplasms of the brain: A clinicopathological study in a tertiary care cancer centre. Neurol Ind 2018;66:733-38.